Substitution of nature’s services with technological alternatives has been pursued with almost religious zeal as societies have industrialized over the past three centuries. But the time for reverse substitution may be upon us. In a wide variety of settings, from water purification to climate change adaptation, investors are increasingly considering the worthiness of green infrastructure solutions, such as mangrove restoration, rather than conventional gray investments, such as sea walls, to achieve the same environmental quality outcomes. But in times of fiscal austerity, cost-effectiveness is paramount. The problem is that infrastructure investors do not have a consistent and robust way to compare gray with green infrastructure in an apples-to-apples manner that is convincing to budget hawks. In addition, uncertainty is greater with “unproven” green infrastructure approaches. As a result, green solutions are often neglected. Here, we present the contours of a general methodology called green-gray analysis (GGA) and demonstrate its usefulness in a green-gray trade-off facing the Portland Water District in Maine. Results provide evidence for the superiority of green investments in several scenarios, purely on financial terms. When ancillary benefits, such as carbon sequestration or passive-use values for Atlantic salmon are factored in, the case becomes even more compelling. A replicable GGA methodology can be one important solution for scaling up green infrastructure investments worldwide.
Investments in “green” infrastructure solutions, such as agricultural and forestry best-management practices, are increasingly recognized as cost-effective ways to achieve environmental quality outcomes over a long time-horizon relative to traditional investments in “gray” infrastructure, such as wastewater treatment or water filtration plants.
Despite this, there is no consistent and accessible methodology available to decision makers for investigating the financial trade-offs.
This paper suggests such a methodology—green-gray analysis (GGA)—and demonstrates its application by considering a green-gray case study involving the Portland Water District (PWD) in Maine.
This application suggests that investing in five green infrastructure options could represent either a cost savings of up to 71 percent over constructing a new filtration plant or a cost increase of up to 44 percent. Uncertainty over green infrastructure efficacy and costs accounted for this wide range of outcomes in six modeled scenarios. The PWD analysis demonstrates the usefulness of a general methodology, but also identifies significant data gaps that need to be filled to make GGA more widely applicable.
By working out the details of a replicable methodology, it is our hope that infrastructure investors who typically opt for gray can routinely evaluate the financial and economic benefits of green in their formal decision-making process.
For almost a century, New York City has drawn its drinking water from the Catskill Mountains, more than 100 miles to the north. In April of 2007, the Environmental Protection Agency (EPA) announced the results of a several-year review of the city’s ongoing program to maintain clean drinking water supplies with forest and open space conservation in the Catskills rather than the construction of filtration plants. The results were encouraging. The EPA concluded that as long as the city agreed to set aside $300 million over the next 10 years to acquire land and restrain upstate development that causes runoff and pollution, the agency would exempt New York from having to build an $8 billion filtration plant.1 The Catskills aqueduct has been held up as the quintessential example of green infrastructure trumping gray and has prompted cities worldwide to consider alternative solutions to the infrastructure demands of the twenty-first century.
Green infrastructure is increasingly recognized as a superior investment. Cities around the country are starting to realize the economic—to say nothing of environmental—benefits of this shifting reality. A recent analysis by New York City found that green roofs and bioswales could help meet water-quality goals with savings of more than $1 billion compared to conventional infrastructure; the Chesapeake Bay could reduce nitrogen loading at less than half the price by using cover crops instead of upgraded wastewater plants. The City of Philadelphia found that the net present value of green infrastructure for storm-water control ranged from $1.94 to $4.45 billion, while gray infrastructure benefits ranged from only $0.06 to $0.14 billion over a 40-year period.2 And using a system of wetlands, North Carolina could minimize storm-water runoff for 47 cents per thousand gallons treated. Using conventional storm-water controls, this figure jumps to $3.24 per thousand gallons.3-5
An emerging hypothesis in environmental management settings is that investment in ecosystem-based green infrastructure solutions provides economically superior environmental quality outcomes when compared to investments in technology-based or “gray” infrastructure. As noted by economists Lucy Emerton and Elroy Bos, “It is increasingly apparent that investment in ecosystems now can safeguard profits in the future, and save considerable costs.”6
While there is no single definition, current literature suggests that green infrastructure comprises “all natural, semi-natural, and artificial networks of multifunctional ecological systems within, around, and between urban areas, at all spatial scales.”7 Examples of green infrastructure investments include reforestation, installation of grass and riparian buffers, green roofs, porous pavement, urban trees, constructed wetlands, stream restoration, and best-management practices for agriculture and forestry. Many of these assets even appreciate in value over time. For instance, as urban trees mature, they do better at cooling down the urban heat-island effect. As riparian buffers mature, they do better at regulating water quality and flow, and supporting fisheries. This stands in stark contrast to gray infrastructure, which only depreciates.
Yet there is no consistent and accessible methodology to compare green with gray. Although existing case studies are intriguing, green-gray analysis (GGA) is in its infancy and has yet to permeate public infrastructure investment decisions. While calculating the costs and benefits of gray infrastructure is relatively straightforward, placing analysis of green infrastructure costs and benefits on equal footing has not yet been formalized. As a result, financial savings potentially gained from green infrastructure investments are not considered or realized. Additionally, ancillary ecosystem service benefits are largely excluded from investment and infrastructure decisions. This methodological gap presents a formidable barrier to public infrastructure investment managers’ contemplating investment in green rather than gray infrastructure.8
But the tools exist. Despite the lack of formal guidance on comparing green with gray infrastructure, the general contour of a standardized GGA methodology emerges from a mix of economic decision theory, portfolio theory, public investment theory, non-market valuation, and a growing list of applied case studies. For example, federal analysts considering the best alternatives for configuring ports, highways, dams, and other major public infrastructure investments are already well steeped in social benefit-cost analysis (BCA) where long-term returns to society in the form of improved navigation, flood control, or transportation savings are compared with capital, maintenance, and operations costs. Green infrastructure presents another alternative that is typically overlooked, but which can nonetheless be compared—albeit with some added complexity—to these more conventional alternatives using standard methods of BCA. Drawing on these tools, we were able to distill six key components for an effective general methodology, which we then tested in Portland, Maine:
First, the investment objective and constraints must be very clearly specified. Although intuitive, this step is essential for getting the math right, which involves making sure all relevant benefits and costs are included in the right units, and in the right places in the investment trade-off equations. Small computational errors—such as neglecting to annualize capital investments over the proper length of time—can have huge implications for the feasibility of either green or gray options. Specifying the investment objective in mathematical terms minimizes the probability of such errors occurring. GGA is a special case of public investment limited to situations where environmental outcomes are sought and where both gray and green infrastructure options exist. A survey of existing case studies and potential applications suggests that most green-gray analyses fall into one of three distinct investment objectives. Each of these objectives can be expressed by mathematical formulas familiar to investment managers and used for selecting the optimal investment portfolio when a diversity of options exists.9
Three Common Investment Objectives for Considering Green Infrastructure Alternatives
- Minimizing the costs of mitigation plus the expected value of losses from natural and human disturbances. Take the example of coastal flooding from storm surge events. Relevant gray investments may include sea walls, dikes, levees, pumping facilities, and floodways. Relevant green infrastructure investments may include restoring mangroves, dunes, and wetlands. In these and other disaster-risk cases, the public investment objective is to reduce the expected value of the loss, which is simply the probability of a disaster or disturbance occurring multiplied by the value of the loss should the event occur.
- Minimizing the cost of meeting a regulatory or planning objective. Reducing the delivery of nitrogen and phosphorous pollution entering receiving water to a target specified by a wastewater treatment plant’s Clean Water Act permit is one example. The nutrient load target can be achieved by upgrading plant technology or capacity, or by financing best management practices on farms upstream.
- Maximizing the net public benefits of infrastructure investments. An example is a fisheries management agency seeking ways to enhance high-value recreational fish resources through either investment in hatcheries (gray) or dam removal and stream restoration (green). Importantly, under this objective, all categories of benefits and costs apply, including both market and nonmarket benefits that may be ancillary to the primary infrastructure investment rationale. These ancillary benefits represent an important component in many existing GGA applications and are critical to efficient investment in environmental management. In particular, investment decisions made on the basis of cost alone undervalue additional ecosystem service benefits produced by green infrastructure and hence may lead to suboptimal investment decisions.
Second, portfolios that include both green and gray investments must be developed. In each infrastructure investment situation, there may be one or more gray, and likely several green, investment options under consideration. Developing a portfolio of gray options is fairly straightforward, as technologies are relatively well understood. There is less familiarity with green options, although the literature on the number and applicability of various green infrastructure solutions is rapidly evolving.10 In constructing green portfolios, there are several unique aspects to consider, such as physical constraints (e.g., there are only so many streams where riparian buffers could be restored); the need to incorporate redundancy (e.g., replanting two acres of trees instead of one in case one burns down); and sequencing (e.g., obtaining water rights before constructing wetlands). A more complete GGA methodology would provide practical guidance on all of the unique aspects of green portfolio design.
Third, the outcomes must be clearly modeled. This is perhaps the trickiest aspect of GGA. Quantitatively establishing the relationship between the level of investment in any one green infrastructure project and the environmental outcomes requires careful modeling that relates changes in ecosystem function to changes in economic services provided. As one example, there has been considerable work relating wetland acreage to associated maximum storm surge heights, and associated losses.11 On the other, there has been almost no work at all relating installation of green storm water controls to reductions in nitrogen or phosphorous runoff—most studies have focused on volume reductions. In these situations, the trickiest step is making assumptions about the relationship when there is little scientific information to go on.
Fourth, present-value costs and benefits of individual green and gray measures must be quantified. Before portfolios of green or gray options are considered, each individual component needs to be analyzed by itself. An important consideration in GGA is to ensure that both green and gray options are analyzed on a common platform so that costs and benefits can be directly compared or combined. The gray infrastructure option should serve as the baseline since GGA is often considered in contrast to some impending gray investment decision, and not vice versa. Adopting gray as the baseline requires evaluation of green options within the general analytical framework offered by standard infrastructure investment methods. The U.S. Environmental Protection Agency provides a useful synopsis of standard two-stage discounting to evaluate gray infrastructure investments.12 Present-value costs and benefits of green infrastructure can be modeled in precisely the same way, albeit with a few complexities.
Fifth, and at the heart of GGA, is using alternative investment analysis to compare green against gray, or different combinations of green and gray together. Once the benefits and costs of individual green or gray measures are calculated, the next step is then to compare full investment portfolios. Depending on the investment objective, this comparison is carried out by using either benefit-cost analysis (BCA) or cost-effectiveness analysis. BCA is a technique that is used to estimate and sum up the future flows of benefits and costs given particular resource allocation or policy decisions. Based on this sum, the value of a particular choice can be compared against some set of alternatives. Cost-effectiveness analysis on the other hand is a technique for identifying the least-cost option for meeting a specific physical outcome.
Finally, one must account for risk and uncertainty. Many believe that green infrastructure investments are generally riskier and more uncertain than gray. But gray is also subject to significant risks, such as plans for maintenance that are never carried out and technological failures that occur all too frequently. Nonetheless, sources of risk unique to green include the possibilities of floods, fires, insect outbreaks, extreme drought, and climate change to significantly affect the function of green infrastructure elements over the long run. Sources of uncertainty include poor existing data on implementation costs, speculative relationships between green infrastructure elements and the environmental outcome sought, and lack of understanding about important land-use trends, market trends, landowner behavior, or policy changes that have bearing on the investment decision. Risk and uncertainty can be dealt with in two fundamental ways—through project design and through project analysis. Redundancy, or having two or more green infrastructure elements included to achieve the same outcome, is one way to reduce risk and uncertainty in the design of green infrastructure investment portfolios. With respect to analysis, standard approaches for incorporating risk and uncertainty include sensitivity analysis, scenarios, and use of expected values.
Portland Water District Case Study
Drawing on these six components, we conducted a case study of the Sebago Lake Watershed in the Portland Water District in Maine (PWD). Sebago Lake contains some of the cleanest water in the Northeastern United States and is also the primary drinking-water source for PWD, supplying drinking water to over 200,000 people daily. PWD currently qualifies for filtration avoidance under the U.S. Environmental Protection Agency 1989 Surface Water Treatment Rule (SWTR). The rule waives public water systems from requirements to install filtration systems as long as concentrations of turbidity and either fecal or total coliform are maintained at or below regulatory baselines through upstream land-use management practices.
In recent years, concerns have been expressed that upstream development, deforestation, and population growth trends may jeopardize the filtration waiver and force PWD to install a conventional filtration system. For example, the U.S. Forest Service has determined that certain areas of the Sebago Lake watershed are at high risk of forest conversion due to development pressure, which, coupled with unsustainable land use practices, are a major threat to water quality.,sup>13 In response, commercial water users, residents, nongovernmental organizations, recreation interests, and other stakeholders are actively investigating green infrastructure alternatives that would minimize the chance of losing the waiver and otherwise help reduce PWD’s water treatment costs. Using the GGA methods discussed above, we completed a preliminary analysis to provide an initial sense of the economic trade-offs involved and to identify the various data gaps and parameters that would need to be addressed for a more complete analysis.
Using cost-effectiveness analysis for our framework, we compared the costs of a new filtration plant with investment in six green infrastructure elements over the next 20 years that together would help maintain the watershed’s high-quality waters. These included riparian buffers, upgrades to culverts, sustainability certification of future timber harvests, reforestation, and conservation easements. The quantity available and costs associated with the green infrastructure portfolio were determined through on-site consultations throughout the watershed, review of publicly available data, and GIS analysis.
We ran six scenarios that represented different levels of investment, analysis periods, assumptions regarding the efficacy of green infrastructure measures, cost assumptions, and discount rates. Our key findings include:
- The present-value life-cycle cost of building a new filtration plant would range from $97 to $155 million across the scenarios.
- The present-value life-cycle cost of green infrastructure solutions range from $44 to $172 million across the scenarios.
- Under four scenarios (see box 2, “The Six Scenarios”), green infrastructure represented a cost savings, with the most optimistic case of $111 million saved over 20 years.
- Under two scenarios (S3, S6) gray infrastructure proved superior. Under the least optimistic for green, green infrastructure would represent a 44 percent increase in costs.
- Uncertainty over green infrastructure efficacy and costs are what accounted for the wide range of outcomes across the modeled scenarios.
- Ancillary benefits in the form of carbon sequestration and Atlantic salmon habitat would make an even more compelling case for investment in green infrastructure. By combining empirical data on the ground with calibrated nonmarket benefits transferred from other settings, we estimate that these nonmarket benefits would amount to $72 to $125 million over a 20-year timeframe. Including these ancillary benefits would make green infrastructure superior in all six scenarios.
Towards More Widespread and Robust Applications of GGA
Three insights emerge from our PWD case study. First, while there are certainly complexities involved, green infrastructure investments can be presented in a manner commensurate with conventional gray investments; the two can indeed be compared dollar-for-dollar, apples-to-apples, by public investment analysts. This suggests that, once fully developed, a GGA methodology could be a standard part of infrastructure investment decisions for a wide variety of settings. Second, the key to an actionable GGA is a robust underlying model that establishes the quantitative relationship between each green infrastructure component and the outcome sought, whether it be regulating pollutants at a specified threshold or generating net public benefits. These models will become more refined and accurate as GGA applications proliferate.
Third, green infrastructure presents additional sources of risk and uncertainty relative to gray. Thus, any GGA must place a heavy emphasis on identifying and mitigating risk and uncertainty through portfolio design (e.g., a green multi-barrier approach), analytical adjustments (e.g., modeling the risk of failure), or sensitivity analysis. Nevertheless, and as demonstrated by the PWD case study, green infrastructure may represent savings large enough to warrant selection, even under conditions of significant uncertainty. A standardized GGA methodology that incorporates accurate cost-estimates and site-specific biophysical models will help investment analysts make the case for green infrastructure, even to the most skeptical budget hawks.
The Six Scenarios
Scenario 1 (S1) is the baseline, and relies on mean infrastructure costs, a discount rate of 3 percent, an opportunity cost of capital (OCC) of 5 percent, a 20-year analysis period, and the residual risk of a waiver loss for green infrastructure set at zero. Green infrastructure elements are implemented up to their maximum level. Scenario 2 presents the best case for green infrastructure by incorporating the lower-bound cost estimates for green infrastructure and the upper-bound estimate for gray. OCC is set at 7 percent. Scenario 3 presents the best case for gray infrastructure by incorporating the lower-bound cost estimates for the membrane filtration system, and the upper-bound cost estimates for green infrastructure. The green infrastructure mix was adjusted to maximize conservation easements, an adjustment that increases the overall cost of the portfolio. This scenario also incorporates risk and uncertainty with respect to green infrastructure by modeling a residual risk of the waiver loss of 25 percent.Scenarios 4, 5, and 6 are variants of the first three. S4 was designed to test the sensitivity of the baseline analysis to relatively minor changes in the discount rate (increased to 4 percent), OCC (increased to 7 percent), and the analysis period (increased to 30 years). S5 tempers the best-case scenario for green (S2) by increasing the risk of waiver loss from zero to 10 percent and by lowering the cost of capital to 5 percent. S6 tempers the best-case scenario for gray (S3) by eliminating the residual waiver loss risk. While the Excel-based calculator used to model the scenarios is capable of analyzing a nearly limitless number of scenarios, we believe these six represent a range of outcomes that are likely to be found in a more complete GGA when more precise data becomes available.
Our partner, the Manomet Center for Conservation Sciences, provided critical information for the analysis of the Sebago Lake Watershed.